Sleep deficiency, poor sleep and night shift work increase risk of cardiovascular disease, type 2 diabetes, obesity, mood disorders and all cause mortality. Sleep disorders themselves pose a large public health and economic burden. Although sleep is a fundamental behavior with a significant genetic contribution, the genetic basis of variability in sleep regulation in the human population and shared biological pathways with chronic disease is almost completely unknown. Sleep duration, timing and quality are heritable, providing opportunities to identify underlying genes and biological pathways. However, sleep phenotypes also depend on social and environmental factors and disease conditions, requiring large datasets and careful consideration of covariates to detect genetic effects. We hypothesize that meta-analysis of genome-wide association studies (GWAS) using existing large-scale publicly available population-based datasets and enhanced statistical methods for admixture association and covariate modeling will identify new genes and biological pathways important for sleep regulation. In order to test this hypothesis, we propose the following specific aims: 1) To harmonize self-reported sleep duration, timing and quality phenotypes across publicly available datasets, and 2) To identify genetic variants associated with heritable sleep traits by performing GWAS and meta-analyses in subjects of European and African American (AA) ancestry. Identifying genes for sleep phenotypes using secondary analysis of publicly available GWAS cohorts is a cost-effective and efficient way to gain insights into biological pathways underlying sleep regulation. This knowledge is necessary for development of novel diagnostics and therapeutics for sleep disorders and for understanding causal relationships between sleep and associated chronic diseases to enable effective interventions for these conditions.